The surgery was only part of the cost, of course. Karen would need help taking care of her daughters after the operation, and she still needed help covering food and other expenses. I told her I would find a way to cover it.

Around this time, I had something of an epiphany regarding the use of AI tools in publishing. I’d been playing around with using AI for brainstorming, sketching scenes, cover design, and a few other tasks, but now it was becoming clear to me that with recent improvements in LLMs and image generation models, AI was going to revolutionize publishing—and I was either going to be on top of the wave or under it. I decided to launch a suite of AI-enabled publishing tools, developed with ChatGPT.

Having no money and nothing else to do while I was in Peru, I spent all my waking time either writing or building apps. My plan was to build the apps primarily for my own use, but to eventually release more polished versions under a subscription model. After spending most of the day trying to come up with a catchy and appropriate name, I said screw it and called it Soda Badger. Long-time readers of this Substack will recognize the reference. (I’ve had to put the apps aside to work on BasedCon and the last Ransom’s Law book, but more will be coming soon, God willing.)

Read more

To back up slightly: while I was still in Colombia, during one particularly bad bout of loneliness in Peru, I made the mistake of unblocking Karen. I say “mistake” because I’m certain that by this point anyone who has been following this story from the beginning is actively grimacing right now. WHY? you may ask. I plead insanity. Some combination of loneliness, sunk cost fallacy, and residual feelings for her made me want some kind of satisfactory resolution, if not reconciliation. I asked how she was doing.

She was, predictably, not doing well. Still unable to work because of her health issues, she had borrowed $1500 from a man who evidently had romantic intentions toward her. As far as I could tell, this man—whom Karen called el imbécil—loaned her the money thinking he was going to be compensated in the way imbéciles usually want to be compensated. When Karen balked at that, he demanded the money back.

Read more

Having spent every dime I had on medical issues—mine and Karen’s—I began looking for an inexpensive place to stay for the next several months. I couldn’t legally return to Ecuador, having hit the six-month tourist stay limit, and now that Karen had broken up with me, I didn’t have much reason to be there anyway. The obvious choice was Colombia: it’s about as cheap as Ecuador, and it’s a short flight from South Florida. I had visited Colombia before, but Cartagena during the high tourism season was above my budget of virtually nothing. Eventually I found a very inexpensive Airbnb in a tiny seaside town called Taganga.

Read more

I managed to get ahold of a friend of Karen’s, named Aleska, and asked her if she knew what was going on. She told me Karen hadn’t been responding to her messages either. I told her I was going to go to Karen’s family’s house and see what was going on. She must have told Karen that, because Karen did finally message me.

Karen said she was having a really rough time, but she did want to see me. She asked if it was OK if she came over with Aleska and Aleska’s (American) boyfriend. I had rented an apartment with a swimming pool (Airbnbs are cheap in Ecuador, and I knew Karen’s daughters liked to swim). Karen said she thought it would be fun if we could all hang out and have some beers. “This is what I need right now,” she said, preempting my objection that I was hoping she and I could have some time alone together. I acquiesced.

Karen came over with her two daughters, Allie and Kaely, Aleska, and Aleska’s boyfriend. We had a good time, and I was particularly happy to see Karen’s daughters, whom I had come to care about very much. But it was clear that Karen was still in pain from her ongoing health issues—and it was also clear that she was avoiding being alone with me. After a few hours, everybody left, and I was again alone.

Over the next several days, Karen kept assuring me she would come over again soon, but she never did. She was minimally communicative and finally told me that she was sorry, but she just couldn’t be in a relationship right now.

She insisted the reason was her poor health and mental state and the toll that her health problems were taking on me. This, of course, sounded like typical female “It’s not you, it’s me” breakup language. The strange thing, though, was that she continued to insist all this time that she loved me very much, that I was the love of her life, that I was the best man she had ever known.

Read more

Read more

In February, I returned to Michigan and took stock of my financial situation. I had been heartened by messages from my marketing consultant, Elaine Ash, telling me she was getting excellent results from the Facebook ads we had been running for the Rex Nihilo books. Now that I actually checked the data, though, I could see that although people were clearly clicking on the ads, they weren’t buying my books. At all. Out of hundreds of clicks I was getting literally zero sales. (In retrospect, I think it was probably bots “clicking” on the ads.)

Additionally, because I had wanted to get a good idea of baseline sales without doing any marketing, I had stopped doing the periodic promotions that I usually did. So not only was I making no money on the ads; I was making virtually no money, period. I had several months in a row where I made around $500 on my entire catalog. I don’t blame Elaine for this; I think she found a strategy that worked, and then something changed and it stopped working. She very graciously offered to return the money I’d paid her. Knowing she’d done a lot of work, I accepted a refund of half the fee.

Meanwhile, the medical emergencies continued. On March 13, 2025 Karen was diagnosed with dengue and hospitalized for two weeks. On April 2, she went in for surgery for ovarian cysts.

My credit cards were again maxed out, and the people in charge of the pension fund were now telling me that I wouldn’t get the money until August. I was about to launch a Kickstarter for a new series, Ransom’s Law, but I wouldn’t get that money until August either. I was by this time thoroughly defeated. It seemed like no matter what I did, no matter how many times I got up, I would always get knocked down again. I was barely keeping it together. During one of the more stressful moments, I checked my blood pressure and it was 180 over 100.

Read more

By this point, I was beginning to wonder whether I were the best thing that had ever happened to Karen or the worst. How was it possible for one family to have so many ailments and accidents? Was I being scammed? I repeatedly went over the evidence, and I couldn’t see how all these events could have been fabricated. Karen had sent me videos from the clinic, X-rays, and pictures of her nephew in a hospital bed with a nurse inspecting his arm. The accounts she sent me were detailed and consistent with events unfolding over time. She told me about how the break in her nephew’s arm had happened and what the doctors were telling the family. She told me about her sister crying in the waiting room.

In other words, if these events were fabricated, I was dealing with an incredibly sophisticated group of sociopathic grifters—people who could probably make vastly more money selling drugs or even running a legitimate business, rather than trying to squeeze a few thousand dollars out of a struggling novelist. Not to mention that I had gotten to know Karen pretty well by this time, and I just didn’t see her participating in such a scam. Being 3,000 miles away, I had no way of knowing for certain, but the way I figured it, the worst thing that could happen to me was bankruptcy. The worst thing that could happen to Karen’s nephew was being crippled and in pain for the rest of his life.

Read more

I started with my first divorce and then moved forward until the narrative got uncomfortably close to the present day, then jumped back in time to my childhood. I moved forward again until I reached my starting point—where my life began to fall apart. The only thing that remains is to relate the events of the past year.

I had hoped that by the time I reached this point I would be able to relate them coherently and honestly—preferably without having a mental breakdown somewhere in South America. I’m still not 100% sure I’m there yet, but I’ve decided to go ahead anyway. By the time I’m done, I think you will understand both my reticence and my decision to forge ahead.

Because of the convoluted way I’ve told my story, I wanted to start with a recap of narrative so far. Feel free to skip the rest of this post if you already know this part, but it does provide some important context for the stuff that comes later.

As a child, I was recognized early as unusually intelligent and imaginative, but I was also absent-minded, socially awkward, and completely unsuited to repetitive tasks like schoolwork. I understood concepts quickly but lost assignments, forgot deadlines, and earned grades that bore little relation to my abilities. My father responded to these failures with criticism and ridicule, frequently treating me as lazy, irresponsible, or stupid. I learned to assume that whatever potential other people saw in me would remain unrealized. I struggled with deep depression and thoughts of hopelessness and suicide.

That pattern continued into adulthood. I barely graduated from Calvin College with a philosophy degree and spent the next several years cycling through menial jobs for which I was temperamentally ill-suited. I was bad at following routines, dealing with customers, and pretending to respect petty authority. I was repeatedly fired or quit in frustration. I had vague ambitions to become a writer, but nothing in my early academic or employment history suggested that I would ever build a stable career, much less become successful at anything.

I married my college girlfriend Julia in 1993, and we moved to central California. I eventually talked my way into a technical-support job in the East San Francisco Bay area despite having little formal computer knowledge. My ability to understand systems and solve unusual problems allowed me to build a career in software. Julia and I had two children, Willem and Madeline. I worked constantly, bought property, renovated and built houses, and eventually started writing novels while still working more than sixty hours a week.

The contrast with my earlier life was dramatic. The absent-minded kid who had been called feeble-minded and the young man who could not keep a grocery-store job was now earning a professional salary, supporting a family, designing and building a house, and developing complex software. I had finally become conventionally successful. But I remained driven by the fear that everything could collapse if I stopped moving, and the same anxiety that powered my accomplishments made it difficult for me to slow down, listen, or share control.

From my perspective, I was doing everything necessary to provide for my family. From Julia’s perspective, I was making major decisions and dragging her along without treating her as an equal partner. She became passive and resentful; I became angry that she was not helping. Neither of us understood the other, and our marriage deteriorated through years of poor communication and accumulating grievances.
Meanwhile, my writing career showed signs of promise. I self-published Mercury Falls, which sold well enough to attract Amazon Publishing. They re-published Mercury Falls under their imprint and then published two sequels. I began to believe I might be able to leave software development behind. Julia seemed almost resentful of my success, so I had no one to share it with.

Eventually Julia maneuvered me into raising the subject of divorce, although I later understood that she probably wanted me to reject the idea and prove that I was committed to her. Instead, exhausted and lonely, I agreed. I felt immediate relief, which only made her angrier. Looking back, I accept that whatever her failures were, I was not strong or emotionally capable enough to hold the marriage together.

Read more

While I was still working at Global Inventures, I started a blog called Mattress Police, where I wrote humorous essays and commentary under the name Diesel. My rantings soon garnered me a small following of devoted readers, and I began to think about more ambitious writing projects. Around this time, I was also selected to be a deacon in our church, and the juxtaposition of my online persona and my very serious responsibilities at my church prompted me to imagine a basically well-intentioned angel who was an incorrigible smartass. The title Mercury Falls popped into my head, unbidden.

I knew when I was writing Mercury Falls that it was going to be a success. That is, I knew that it would resonate with some segment of the population. I wasn’t sure whether that segment was ten thousand people or ten million people, but I knew that making Mercury Falls into a success was simply a matter of finding those people. I was not, however, under the impression (yet) that I could make a living writing.

A few months into my “retirement” from Global Inventures, when Mercury Falls was maybe half written, I got a call from another ex-GI employee who asked me if I’d like to do some work for Google.

My friend worked for a company that had a contract to handle Google’s internal inventory. Basically, this company was charged with making sure all of Google’s roughly 20,000 employees have all the laptops, monitors, mice, keyboards, and other crap they need to do their jobs. Somewhat hilariously, they kept track of all this stuff with Excel spreadsheets. My job would be to come up with a better system.

At this point we were still flush with cash from our refinance, and I had underestimated how much Stage Two of our house was going to cost. As a result, I was under the impression that I didn’t really need a job. Still, I thought it might be fun to work for Google, so I agreed to drive in to Mountain View for an interview. If I had any idea that I’d be burning through the remainder of our savings over the next six months of home-building, I might have been nervous, but due to my misplaced confidence, I was completely nonchalant. I may, in fact, have been a little flippant. I remember at one point explaining to the interviewer—apropos of nothing—that a cougar, mountain lion, and puma are all essentially the same animal. I think I figured that they might as well get a dose of my personality early on, so there wouldn’t be any surprises later.

Evidently my glibness came across as confidence, because my friend called me a few hours after the interview to tell me that the Google people loved me. I’m pretty sure I responded by laughing out loud. In any case, the next week I started working on a system to track Google’s inventory. I called the system “Warewolf.”

I’d like to say that Warewolf revolutionized the way that Google handles inventory and saved it hundreds of millions of dollars (and honestly, it could have), but ultimately Google is like any gigantic company: projects are canceled all the time for political reasons far beyond the control of a peon like me. One week Warewolf was on the verge of being rolled out to all of Google’s locations, worldwide, and the next week it had been targeted for termination. The last I heard, Google was months overdue on rolling out a much more expensive system. I will not comment on the rumor that certain hardware I borrowed from Google was never returned because I knew for a fact they had no way of knowing I had it.

After Warewolf was canceled, I realized that, ironically, the output I produced at my “real job,” for which I was well-paid, was actually more fleeting and ephemeral than the frivolous creative work I did in my spare time. If I was going to leave any sort of lasting mark on the world, it was going to be in the form of something I had written, not some wonderful new software application.

Working at Google, I was making enough money to afford to pay a contractor to finish Stage Two of our house. (Having spent a few thousand hours working on Stage One, I had come to the conclusion that I didn’t really like building houses, and would rather pay someone to do it for me – especially if the person I was paying could do it better and faster.) This gave me time to finish Mercury Falls. Once I was happy with the book, I began to shop it around to agents. A few of the agents I contacted asked to see the full manuscript, but in the end I couldn’t interest any of them in representing it to a publisher and decided to self-publish it.

I gave Mattress Police readers a month to pre-order a signed copy of the book, and by the release date I had received about 100 orders. While the pre-orders were coming in, I spent some time scouring Amazon.com for the email addresses of people who had reviewed (and liked) similar books. I sent out over 100 review copies, which I paid for with the money I had made on the pre-orders.

Mercury Falls sold another dozen or so copies on Amazon.com over the next two weeks. Then I published an electronic version for Amazon’s Kindle reading device. Initially, I priced the Kindle version at $4.99, and over the next two weeks I sold another twenty copies or so. Then, as an experiment, I tried dropping the price to $2.99. To my surprise, the rate of sales more than doubled, so that I was actually making more money at the lower price. After two weeks of that, I tried dropping the price to $0.99. Sales more than doubled again. Not only was I making more money at $0.99 than I was at $4.99, I was reaching several times as many readers. I couldn’t figure out why more authors didn’t price their books at $0.99. Over the next six months, Mercury Falls sold over 4,000 copies—making it more successful than most first novels published by traditional publishers.

The success of Mercury Falls caught the attention of Amazon.com’s new publishing venture, AmazonEncore. They offered to re-release Mercury Falls as an AmazonEncore title, with the full weight of Amazon.com behind it. I accepted. The new edition of Mercury Falls was published almost exactly a year after I originally self-published it. Over the course of the next several months, Mercury Falls sold another fifteen thousand copies. For two days, it was actually the bestselling Kindle book on Amazon. Clearly Amazon had found some of those potential fans I knew were out there. I started working on the sequel.

I continued to do software development on a contract basis, while writing novels in my spare time. Thanks to legacy publishers’ failure to capitalize on the growing popularity of Kindle—and a massive investment by Amazon to become a big player in the publishing industry—for a couple years I actually made more writing novels than writing software. I was doing so well, in fact, that after a particularly nightmarish contracting job I decided just to write full time. This was, it turned out, not the smartest move.

And that’s about where I started with I began posting to this Substack a little over a year ago. Not long after leaving my final contracting job, my entire life fell apart. I lost my house, got divorced, moved 3,000 miles away from my kids, and nearly went bankrupt. Things got better after that, but only enough for me to lose everything again. In fact, I started this Substack in part to work through some of the disastrous and traumatic events of the previous thirteen years, in the hopes that the worst of it was over.

Friends, it was not.

In some ways, the past year has been the hardest one so far. I retreated to posts about my childhood because I simply didn’t have the perspective on recent events to relate them with any kind of clarity. Well, I’ve run out of stories from the distant past, so I suppose it’s time to catch you up on the latest series of disasters.

Because I’m not ready to publicly post many of the events of the past year, most of the upcoming posts will be protected. If you’d like to upgrade to a paid subscription to read them, it would be much appreciated.

Stay tuned.

Unfortunately, Allie is once again running a fever and has had to be hospitalized again. Needless to say, Karen still does not have the money for this. Karen is suffering from some serious health problems herself, and it is extremely taxing to have to deal with this on top of her own problems. I’ve been trying to help Karen’s health and finances stabilized to the point where she can work, but it’s just been one crisis after another for close to two years now.

The GiveSendGo campaign is still active, if you feel moved to help. Sharing this post is also helpful.

Thank you.

Donate to help Allie

Why did I suddenly pivot from writing to building apps? Mainly because AI models have advanced to the point where—with a little guidance—they can perform a wide variety of publishing tasks extremely well, allowing authors to save money, market their books more effectively, spend more time writing, and produce better books faster.

But the other reason is that AI models like ChatGPT can now write code much faster (and usually better, if given the proper guidance) than a human being. That means that someone with a software background and a clear idea of what the app needs to do can build an app about ten times as fast as they could a few years ago. So fast, in fact, that I’ve started building all sorts of little apps for tasks where in the past I might have had to hunt for an app online or just created yet another Excel spreadsheet.

A few weeks ago, while playing an online version of the classic boardgame Acquire, my friends and I were getting annoyed with some of the quirks of the software. We’ve found a few different versions of the game online, but none of them work particularly well. I joked that I was going to build my own version of Acquire while we were playing the game.

Well, it took me a little longer than the three hours we were online, but I did it.

I used a Ransom’s Law theme because why not?

You can check it Acquire (Ransom’s Law edition) here:

Play Acquire

You’ll need to create a SodaBadger account, and since there’s no single-player version yet, you’ll need a friend or two.

It should run pretty smoothly, but if you run into any issues or have suggestions for improvement, let me know! Keep in mind that I built this in my spare time (ha!) in between writing novels, planning a sci-fi con, and building a suite of publishing tools.

(And don’t worry; Ransom’s Law book 4 is nearly complete. I should be able to send it to Kickstarter supporters in about 3 weeks.)

If productivity were measured in PowerPoint presentations, Human Inertia would have single-handedly skewed the GDP. He had slides for every possible made-up statistic or unlikely hypothetical situation. There was virtually no correlation between anything on the slides and anything that actually existed. There were slides for profitability, customer satisfaction, gnomes, unicorns, etc. I used to imagine his presentations devolving into a impenetrable Mobius strip of self-reference: “Here’s a graph showing the ratio of time I spend doing PowerPoint Presentations to time time spent doing actual work; here’s one showing the percentage of graphs in this presentation containing pure fabricated nonsense; here’s a graph showing the alarming escalation of the the pointlessness of these graphs....”

At one point there was a plan under discussion that involved getting approval from Human Inertia before starting any projects. I was astounded that they would consider such a plan, because it would essentially prevent any work from ever being done. Every proposal would languish in limbo forever, collecting dust on Human Inertia’s desk. I was all ready to raise a big fuss about this new process being completely unworkable when I remembered a crucial fact: before it could be enacted, the approval process itself would have to be approved by Human Inertia. The problem resolved itself. This would not be the last time that I would be spared by the H.I.P. (Human Inertia Paradox), in which the prevention of work is prevented by the prevention of work (I later incorporated this as the Bureaucratic Inertia Paradox in my first novel, Mercury Falls).

The only thing that really worried me about Human Inertia was that he was technically in charge of the company’s finances. I tried not to think about it, but every once in a while he’d say something so dumb that I’d have to immediately run to the bank and cash my paycheck, just to make sure they had real money in their account.

He was the kind of guy who used to repeat idiotic urban legends like the one about how the average person swallows four spiders a year in their sleep. I can’t fathom what, if anything, is going on in the head of someone who repeats something like that. Did he imagine that there was a university somewhere that had greenlit a study in which people were observed for eight hours every night to find out how many spiders they swallowed? I tried to envision what the log book for that study would look like:

Day 1: No spiders

Day 2: No spiders

Day 3: No spiders

Day 4: No spiders

Day 5: No spiders

Day 6: No spiders

Day 7: No spiders

Day 8: No spiders

Day 9: No spiders

Day 10: No spiders

Day 11: No spiders

Day 12: No spiders

Day 13: No spiders

Day 14: No spiders

Day 15: No spiders

Day 16: No spiders

Day 17: SPIDER!!!!

Day 18: No spiders

Day 19: No spiders

Day 20: No spiders

...

What would the interpersonal dynamics of such a study be? Would the observer tell the subject if he swallowed a spider on a given night?

“So, how’d you sleep last night, Bob?”

“Umm, pretty good. Why?”

“Oh, no reason. No reason.” (Stifles laughter)

“Did I... I swallowed a spider last night, didn’t I?!”

“Nooooo.... I mean, I’m not really supposed to say either way, but no, you didn’t.”

“Really?”

“Really. No spiders. None. I haven’t even seen any spiders, to be honest.”

“You promise?”

“I promise.”

“Okay, good. But you would tell me if I did, right?”

“Well, I’m not supposed to, but yeah, I would.”

“Good, thanks. Whew. I feel better.”

“Hey, that’s what I’m here for. So... any weird dreams last night? Like maybe about something crawling into your mouth?”

Human Inertia had almost no attention span. I used to send him long, boring emails with some crucial information that I didn’t want him to know buried in the third paragraph. He’d find out about whatever it was too late to stop it, and I would just shrug and say, “I sent you an email….”

If you really wanted him to know something, you had to break it into three bullet points, each no longer than this sentence. He could, on a good day, digest up to five bullet points, or 1 2/3 emails. He would copy and paste these bullet points onto a slide, which he would then present at the next quarterly meeting, along with a graph indicating that he fully understood roughly 14% of the bullet points he had been sent that quarter. I once joked that I was going to draw a cartoon of him as Moses receiving the Ten Commandments on Mount Sinai. He would be saying to God, “This is great, but can you boil it down to three bullet points?”

As frustrating as it was working for Human Inertia, he was far from being the biggest jerk at Global Inventures. That honor went to a manipulative snake of an account executive (salesperson) who happened to also be named Rob. This is the problem with being a Robert born in 1970: everywhere you go, there is already another Rob. If you’re lucky, he’s not a complete fuckwad.

This Rob, unfortunately, was a complete fuckwad. Whenever a new employee came on board, I would introduce myself as “Rob… but not the bad Rob.” I would add that I wasn’t necessarily saying that there was a bad Rob, but if there was, I wasn’t him. This got tiresome after a while, and eventually a co-worker suggested that they come up with a nickname for me. She made the mistake of asking me what I’d like my nickname to be, and I replied, without hesitation, that I’d always wanted to be called Diesel. To my surprise, the nickname stuck. From that day on, I was Diesel. This just goes to prove that you should always have a nickname for yourself picked out, just in case.

There are as many novel-writing processes as there are novelists, but one thing we have in common is that we all hate writing blurbs. Trying to summarize a book you’ve spent months or years on in a few sentences of marketing speak is a special kind of torture. If AI accomplishes nothing else, it will be worth boiling the oceans* if I never have to write another Amazon blurb.

Give it a try. It’s free!

Try BlurbFixer free

More tools and applications on the way….

*This is a myth. Activists vastly overstate how much water data centers use. Also, water cools down and then you can use it again.

Still, I stuck with The Plan: we moved a mobile home onto the property and moved into it. It was nothing special, but waking up in the middle of an orchard to the sound of birds chirping was certainly an improvement over starting the day in that dank shithole of an apartment. Once the mobile home was in place, we started working on the house. We had designed the house as a two-stage project: first we would build a small two bedroom, one bathroom house, and later on, when we had the money, we would add on three more bedrooms and two bathrooms and convert the existing bedrooms into a garage. At this point, of course, we didn’t even have enough money to finish Stage One, but there was nothing to do but press on and hope things would work out.

I applied for every web-related job I could find, even those well over an hour away, but couldn’t even get an interview. I even applied for teaching jobs. In one particularly desperate moment, I even called up ADP to see if they were hiring tech support representatives. They didn’t call me back.

Then one day in September 2003, after our bank balance had reached zero and we were nearing the limits on our credit cards, I got a call from one of the owners of the consulting company I had worked for after ADP. He was working as the IT director of another small company in San Ramon named Global Inventures, and he wanted to know if I’d be interested in doing some contract work. I said yes.

What was initially supposed to be a three week contract was extended several times, and in 2004 I was hired full time. Eventually I became their “director of application development.” We were able to finish Stage One of our house, refinance the property, and make our balloon payment to the seller.

Working at Global Inventures was an alternately exciting and extremely frustrating experience. GI was a typical tech startup: lots of excitement, long hours, almost no documented processes. The real problem, though, was that this company had been in startup mode for fourteen years. That’s like having your wedding night last for three weeks. I know, it sounds great, but eventually getting no sleep and banging your head against the wall just to demonstrate your commitment gets a little old.

GI only had about thirty employees, but somehow even at this small company the employees had fractured into several distinct cliques – and once again, I was somehow able to get along with just about everyone. Some combination of charm and apathy allowed me to rise above the petty squabbles of the different cliques. I once joked that GI was just like high school, the only difference being that I was a lot better at it the second time around.

It took me a good three months to figure out what it was that the company did, exactly, and I even once I thought I understood it, I found it almost impossible to explain to other people. Here’s my best attempt. Feel free to skip the next four paragraphs if you don’t really give a shit.

You know what a USB connection is, right? USB stands for Universal Serial Bus, and it’s the de facto standard for hooking up crap to computers. Why is USB so popular? Well, in the 1990s there were all kinds of different ways to hook up crap to computers, which became confusing and inconvenient for consumers. So a whole bunch of companies (both computer-making companies and crap-making companies) got together and said, “We need standard way to hook up crap to computers.” They put a bunch of engineers in a room and came up with USB. So now all these companies – many of whom are ordinarily competitors – all make their computers and other crap with USB connections. Crap-making companies don’t have to worry whether their crap will connect to the latest computers, and computer-making companies don’t have to worry whether their computers will connect to the latest crap. And consumers don’t have to have eighteen gazillion different kinds of cables and adapters and doodads to get all their crap to connect to their computers. Everybody is happy.

As you can imagine, though, it isn’t easy to get, say, HP and Apple to agree on something. Or HP and Apple and Sony. Or HP and Apple and Sony and GE. Or all those companies, six other Fortune 500 companies, and a hundred other little companies.

Basically, what you have to do is form an independent organization called a consortium. Each affiliated company sends representatives to speak for that company’s interests in the consortium. A consortium may be just an informal gathering of engineers from a few different companies, or it could be a huge international organization involving hundreds of other people. In other words, these consortia are skeleton companies: they have all the needs of a typical company, but they have no actual employees because everyone who works for them is really working for the company he or she represents.

GI acted as the support personnel for these fictitious companies: we built their websites, answered their phones, ran their marketing campaigns, organized their events, collected their dues, etc. Another way to think of GI is this: we were glorified virtual assistants.

As GI’s chief software developer, it was my job to build software that helped the members of these consortia get work done. “Get work done” is a deliberately vague phrase; on the occasion that I would sit and reflect on what it was that I was doing, I could hardly believe anyone was paying me to do it. Essentially, what I did was build programs that would make it easier for people from different companies to talk about obscure technical rules which, if they were adopted, might someday allow them to build some sort of device that would facilitate the connection of two other devices, which would conceivably save someone some time and/or money at some indeterminate point in the future. The way I figured it, I was at all times at least four levels from doing any actual work. And yet, for reasons I don’t fully understand, it paid a lot better than delivering pizza.

After I became the “director of application development,” my supervisor was a guy I nicknamed Human Inertia because of his uncanny ability to prevent anyone around him from accomplishing work of any kind. GI was a classic case of “too many chiefs and not enough Indians”: the company had 35 employees, of whom seventeen were either a “director” or “vice president.” As a result, it was impossible to do accomplish anything at GI without pissing off some nominal bigwig. On top of this, Human Inertia was irrationally afraid of ruffling the feathers of any of the chiefs, which meant that he was always pooh-poohing my plans to radically alter the way we handled software development.

Every week we would have a team meeting at which I would tell Human Inertia what he had failed to prevent me from doing over the course of the previous week. He would throw some additional obstacles at me, which I would ridicule or ignore, depending on my mood. One time he told me that I wasn’t really allowed to work from home two days a week, to which I responded, “Okay, but I’m going to keep doing it.” I smiled to indicate that it had been a good talk.

To be continued…

The long version:

Like most novelists and other creators, I’ve had mixed feelings about AI.

When it first became clear that AI was going to intrude into creative work, I felt both fear and excitement. Fear, because writing fiction is already a hard way to make a living, and just when I finally felt like I had a handle on the process, along came something that seemed likely to upend everything. Excitement, because before I was a novelist, I was a programmer—and part of me wanted to know what this technology could actually do.

I wasn’t convinced AI could ever write a Rex Nihilo novel. I’m still not. But I knew one thing: if AI was going to become part of the creative process, I wanted to be the one controlling it. I wanted to ride the wave, not get drowned underneath it. So I started experimenting.

After a bit of dabbling, both my fear and excitement faded a bit. The limitations became obvious quickly. AI could be impressive, but it could also forget basic facts, misunderstand instructions, hallucinate answers, or confidently promise things it couldn’t actually do.

But over the next couple of years, two things happened. First, the tools got dramatically better. Second, I started to see that the real opportunity wasn’t in asking AI to “be creative” in some vague, magical way. The opportunity was in understanding what AI is good at, what it is bad at, and how a person with a clear vision can use it to improve specific parts of the creative process.

About two months ago, while working on my Ransom’s Law series, I had something of an epiphany: the main reason people struggle to get good results from AI is that they treat it like a human being. Then they get frustrated when it forgets something obvious, loses track of the project, misunderstands the task, or says it can do something it can’t. The problem is scope.

Give ChatGPT comprehensive instructions and a focused job, and it can write a decent scene. Give it a different set of comprehensive instructions, and it can produce a useful outline. But trying to make the same instance do both at once—and then jump to a different novel, and then remember your branding strategy, and then critique a cover—is where things start to fall apart.

Humans are not as fast as computers at processing information, but we are much better at switching scopes. We know when we are talking about this chapter versus the whole novel, this book versus the series, this character’s motivation versus the author’s marketing problem. AI needs help with that.

That realization led me to start building a set of focused AI tools for writers and creators under the Soda Badger umbrella (why Soda Badger?). The idea is not to replace the writer. It’s to build tools with clear boundaries, clear context, and specific jobs.

(The next step in this process is building apps that act as scope-management tools, so you can take advantage of AI’s speed and power while seamlessly switching from narrow tasks to broad planning, and from one project to another—but that phase is still a few weeks out….)

The first of the focused tools is now live.

CoverFixer is a practical AI tool for authors who want feedback on a book cover. You upload a cover, add a little context about the genre or audience, and it gives you a critique focused on things like readability, composition, title treatment, genre clarity, and overall market fit.

Right now CoverFixer can:

• Review uploaded cover images

• Give practical, market-oriented feedback

• Generate revised cover concepts for logged-in users

• Save recent jobs so you can return to earlier reviews

A pro version with more features and fewer limitations is under development. You can sign up to be notified when it’s released. If you’d like early access to the pro version and other Soda Badger tools, you can also request to be a beta tester.

Sign Up

CoverFixer is not meant to be a replacement for a human cover designer. It’s intended as a quick makeover with an eye toward making the cover more visually striking, legible at a small size, and genre-appropriate. And because the generation feature depends on external AI services, occasional delays or failed generations can happen.

But even with those limitations, I think CoverFixer solves a real problem. Authors often know something about a cover feels off, but they don’t always know what. CoverFixer gives you a fast first pass: not the final answer, but a clearer diagnosis and a direction to explore.

You can try it here:

Try CoverFixer

More tools are coming. Some will focus on writing. Some on planning and organizing. Some on marketing. The common thread is that they will all be built around the same idea: AI works best when it has a specific job, clear context, and a human being firmly in charge.

Stay tuned!

The first thing you need to know is that I’m in a tiny town in Colombia called Taganga—essentially because it’s about the cheapest place in the world for me to stay while I attempt to rebuild my finances after a few disastrous years. “Move to South America for a while” is not a solution that occurs to most financially strapped individuals, but fortunately I speak reasonably good Spanish and can work from anywhere.

This is actually the second time I’ve stayed in Taganga—I started off in Salinas, Ecuador in September, then went to Taganga, then Lima, Peru, and then back to Taganga, with stops in South Florida in between. Part of the reason I came back here is that I developed a good rapport with the couple who runs the Airbnb—and their seven-year-old granddaughter, Valeria, who gave me hugs every day at a time when I really needed a hug. Valeria is with her mom in Medellin right now, but she’s coming to visit in June.

Anyway, when I was in the U.S., planning my second trip to Taganga, I messaged the manager, asking if my room was available in May. She said no, but there were other nice rooms she could give me for the same price. I said that sounded fine; I’d let her know when I arrived.

I arrived, selected a room, and asked if I could pay with a credit card. She said yes, but it would cost five percent more than paying cash. I said OK, I’ll go into town and get cash.

Ordinarily—that is to say, if you’re not me—it’s straightforward to get cash in Latin America. You go to an ATM, agree to pay some ridiculous fee, and get your money.

For me, it’s different. You see, something weird happened the last time I was in Colombia: I used my ATM card all over Cartagena, then traveled to Taganga—and suddenly my ATM card didn’t work. I went to the larger city nearby, Santa Marta, and tried five different ATMs there. Same problem—a vague error message saying the transaction couldn’t be completed. No other explanation.

I called CapitalOne, and they couldn’t figure it out either. Making a wild-assed guess, I asked if it had something to do with CapitalOne buying Discover. After putting me on hold for ten minutes or so, the customer service rep came back and said, basically, “Yeah, your ATM card is a Discover card now. You’ll have to find an ATM that accepts Discover.”

Amigos, there are no ATMs in Latin America that accept Discover.

To recap, my perfectly good ATM card suddenly became an inert piece of plastic in between Cartagena and Taganga because one giant international corporation bought another giant international corporation—and nobody bothered to tell me anything had changed.

So the next time I was in the U.S, I opened another bank account, not affiliated with Discover. This was very smart of me. Then I forgot to put any money in that account before I left for Taganga again, which was less smart.

I explained the problem to the manager as best I could, trying my best not to sound like I was making up an absurd story to avoid paying the rent (“Hay una corporación llamada Discover, y mi banco la compró...”). I had transferred some money into the account, but it would take a few days to arrive. “A few” turned into five, as this happened to be Memorial Day weekend. “Hay un día festivo llamado Día de los Caídos….”

At last the money arrived. I went into town and used my brand new ATM card—which also did not work for some reason. It was not, as far as I knew, a Discover card, but it was apparently equally useless in Colombia. “Lo siento, tengo el dinero, pero los cajeros automáticos no me funcionan….”

At this point, I’d have just given up and used my credit card, but I only have one credit card with a relatively low limit, and I was now at it. I could have made a payment, but I had no idea how long that would take. I had to send myself money by Western Union. A couple fun things about Western Union:

(1) Their website is virtually impossible to use. You can easily blow an hour trying to send yourself $300, and you will still probably fail. With sheer persistence, you may eventually manage to navigate the barrage of timeouts and nonsensical error messages and complete a transaction.

(2) There are Western Union affiliates all over Latin America. Some are corner stores or fruit stands; some are a random person’s house. The one thing they all have in common is that despite proudly displaying a Western Union sign, none of them will ever be able to give you any cash. You will not be able to understand the explanation, but the gist is that you shouldn’t expect them to act like a Western Union branch just because they have a big Western Union sign in their window, gringo estupido.

So I had to take an Uber to the genuine Western Union office in the mall in the next town, where I was at last able to get the hard pesos I needed. I returned to the Airbnb with the cash.

Unfortunately, the dollar had fallen slightly against the Colombian peso since my last visit, which meant I had underestimated the amount I needed by about $50 (or like 80 bajillion pesos). I sheepishly explained that I would go back to Western Union the next day to get the balance (if I could by some miracle get their website to work two days in a row).

As if I hadn’t already damaged my reputation with these people enough, the same day, I somehow managed to lock myself out of my room and had to message the manager asking if she could bring another key.

There was, of course, no other key. Her husband showed up and, after a half an hour of fiddling with a butter knife and a spatula, managed to get the door open. I thanked him profusely and apologized for the trouble. I promised to be more careful with the key.

By this time, the couple were no doubt wondering if I’d been dropped on my head as a child. To my credit, I avoided locking my key in my room again. I did this mostly through the clever strategy of leaving my door open.

Anyway, the next day I was standing in the kitchen, putting sugar in my coffee, while the manager cleaned. Suddenly she exclaimed, “Robert! No puedes mover el aire acondicionado asi!” (“You can’t move the air conditioner like that!”)

My room, you see, had a portable air conditioning unit on wheels, and I would sometimes turn it to point toward me, because being in Taganga in May is like being on the Sun. The unit had a drainage hose that—in typical Latin American fashion—had been stuck through a crude hole somebody had drilled in the wall. I had moved the unit too far and pulled the hose inside, so now there was water all over my floor.

At this point, I did a literal spit take—not because I was surprised, but because what I had put in my coffee was not sugar, but salt.

I hope Valeria gets here soon, because I could really use a hug.

First, Ransom’s Law book two, Adverse Possession, is out today on Kindle!

Second, thanks to your generosity, Karen’s daughter is coming home from the clinic today. Thank you!

The day I started as the webmaster for the city of Tracy, California, I set out (with Stoner’s blessing) to radically revamp the city’s website. I met with various city managers and other key employees to gather requirements for the site and then started building out the site.

Most of the project was what you would call a “back-end” redesign; that is, my job was primarily to fix the underlying structure of the site, not to make it pretty. But I’m also something of a graphics guy, and the site definitely needed a facelift as well. So as part of the project, I also redesigned the look of the site.

As I neared completion on the redesign, I did a demo for Stoner. He was thrilled with the back end changes, but seemed concerned about the look of the site. The problem, he said, was that any significant design changes would have to be approved by the city council.

Ok, I said. So do we have to present it to the council? I’ll be happy to go to the next meeting, if that’s what I need to do. He said he would try to get a demo on the agenda.

A few days later, I asked if he had been able to get the website demo on the agenda. Not this time, he said. Maybe next month. I was a little puzzled by his hesitancy to commit to taking any action, but I figured he just had a lot of other stuff going on.

Around this time I started to become aware of a lot of low grade grumbling about Stoner from the other employees in the IT department. I got the impression that he wasn’t well liked or respected by the people who reported to him. Still, I hadn’t experienced any serious problems myself, so I reserved judgment.

I moved on to another project, meanwhile continuing to occasionally ask Stoner about the status of the council meeting demo. Several weeks passed, with city council approval of the design the only thing holding up the site redesign.

Then one day I was meeting with the city’s director of economic development when she asked what was going on with the web site redesign. When I told her about the holdup with the council, she laughed. “The city council has never had anything to do with approving changes to the website,” she said.

I called another person who attended all the council meetings, and she confirmed what I had been told: it wasn’t the city council’s job to approve design changes, and Stoner knew it. What the hell? Why would he have made up something like that?

I went into Stoner’s office and told him what the economic development director had said. He hemmed and hawed for a bit and then finally admitted the truth.

“I don’t like the design,” he said.

I was rendered nearly speechless. Stoner had stalled the entire web site project for six weeks with a ridiculous lie because he was afraid to tell me he didn’t like my design work. I couldn’t imagine what had been going through his head. What did he think I would do when he told me he didn’t like the design? Burst into tears? Walk out? And how did he think making up a story about needing the approval of the city council was going to solve the problem?

The thing is, design is inherently subjective. Anyone who has done design work has had the experience of creating what he or she thinks is a fantastic design only to have the client pooh-pooh it because they don’t like that shade of green. If Stoner had just said, “The back end is great, but the design blows,” I would have been disappointed for about thirty seconds and then scrapped the design and started over.

But not after six weeks of unnecessary delays. I don’t like being jerked around and lied to. You had your chance to give me your input on the design, I thought. Screw you. This is the design.

“Ok,” I said. “You don’t like the design. Obviously I do like it. But you and I aren’t the end users. We’re not the ones who really matter. Why don’t we send the design out to all the stakeholders in this project and ask for their feedback?”

He could hardly say no to that. So I wrote up a very diplomatic, even-handed email asking for feedback on the design. The response, as I expected, was overwhelmingly favorable. Everybody liked my design better than the old one. The new web site was launched, complete with my design.

Now a smart manager would have been happy with a win for his department, regardless of whose idea the design was. A particularly cagey manager might even have found a way to take credit for the design that he fought against. In retrospect, I don’t think either of those options even occurred to Stoner. He was too busy seething over the fact that I had made a fool out of him in front of his department. They already didn’t respect him, and now this.

Oblivious to Stoner’s fragile mental state, I continued to work on other projects. I had been meeting with the economic development director about the creation of an interactive web-based map that businesses could use to find office space within the city. This project had languished for over a year with little progress because no one involved had the technical expertise to make it work. Now, after three weeks on the project, I had built a rough prototype of the application. I sent an email to Stoner apprising him of the progress and included a link to the development copy of the application so that he could see for himself.

Unknown to me, Stoner set up a demo of the application with his manager, a bigwig with a title like manager of city services or something. Now demo-ing a development copy of an application is iffy at best. Demo-ing a development copy of an application without telling the developer what you’re doing is like driving a random car off the blocks at your mechanic’s shop without telling the mechanic. You’d have to be an imbecile.

Predictably, the application broke during this unplanned demo, and Stoner ended up looking like an idiot in front of his manager. After the meeting, Mr. I’m Afraid to Tell You I Don’t Like The Design stomped up to my desk in a rage, telling me that I had just made a fool of him. “Do you even bother to test your work?” He demanded.

This time I really was speechless. Was he serious? Was it possible for someone to be that stupid? I just sat there, dumbfounded. What did he expect me to do, test the code before I wrote it?

Soon after this outburst, he left for the day. Now as it happened, this was the Wednesday before Thanksgiving, so the office would be closed until Monday. I stayed at my desk late that evening, fuming about Stoner’s absurd tirade. I should have let it go, but I knew that Stoner’s words would be echoing through my head for the next five days if I didn’t do something.

So I shot off an email to him, noting that I had specifically told him that what I had sent him was a link to a development copy of the application and that it was unrealistic to expect an early prototype to work perfectly, especially considering that I was working on the code while he was doing the demo. I brought up the fact that under his leadership the project had gone exactly nowhere for over a year, whereas I had made significant progress in only three weeks. I closed the email with the observation that every failure related to this project had a single point of commonality: him.

In hindsight, it should have been clear to me that Stoner was a psychologically unstable individual. I should have known that my email would set him off. I mean, hell, even a well adjusted person doesn’t like being told he’s an idiot. But I foolishly believed that he was, at some level, a rational individual who would act in his own self-interest. With the amount of recognition that Stoner was getting for my efforts, it would have been insane for him to fire me, no matter how much he personally disliked me. I thought he would read the email and think, “Damn, I’ve pissed this guy off. I need to back off or he’s going to quit.”

But here’s the lesson: Stupidity trumps self-interest. If you have a choice between working for a stupid person or an evil person, pick evil. Evil people are predictable. They won’t screw you unless it does them some good to screw you. Stupid people will screw you because they couldn’t figure out what else to do. And stupid people who are paranoid from smoking way too much pot are even worse.

I came in the Monday after the Thanksgiving weekend to find that I had been locked out of my computer. That’s how I discovered I had been fired. Stoner’s door was closed, and he wouldn’t answer when I knocked. Finally, after an hour of sitting at my desk wondering if I should just pack up and go home, he called me into his office.

“I have to let you go,” he said.

“Why?” I said.

“The official reason is ‘no reason,’” he replied. I was still on a probation period, so he could fire me for ‘no reason.’ Presumably firing me for a reason would have caused all sorts of havoc.

“Would this have anything to do with a certain email I sent last Wednesday?”

“It might,” he admitted.

I eventually gathered from his comments that he thought there was a “conspiracy” against him in his department. He couldn’t easily fire any of the other conspirators, but he could get rid of me.

The sad thing is that he was right: There was a conspiracy against him. Unfortunately for him, the conspiracy consisted of everyone who worked for him except for me. His employees had long since tired of his unstable behavior and capricious management style and had been plotting to get him replaced. As a conscientious employee who did my best to stay out of office politics, I was actually the closest thing that he had to an ally. And he had just fired me.

I’d like to report that Stoner eventually became a victim of his own delusions and got fired. I would have wagered that he’d have been out on his ass within six months. Unfortunately, a Google search revealed that he was still there, five years later.

So I suppose the real moral of this story is that paranoia and rampant stupidity are no barrier to a long and prosperous career in local government. Go figure.

I remember driving home that day and stopping by my son’s preschool because I didn’t feel like being alone. The blow of losing that job was hard to take. Getting the job in the first place had been a fluke; there was no way I was going to find another job like that in the current market.

What the hell was I going to do?

Right now, her younger daughter has a high fever and needs hospital care, but the family doesn’t have the money to cover it. Karen is recovering from surgery herself (her third in two years) and isn’t able to work or fully care for her daughter at the moment.

I told them I would try to help by sharing this.

The goal is to raise $1,000 to cover a few days in the hospital—enough to make sure her daughter gets proper care and to give Karen some relief.

If you’re able to contribute, it would mean a lot. If not, sharing this with others who might be able to help is appreciated.

You can find more details here:

https://www.givesendgo.com/help-pay-for-medical-care-for-a-young-gi

And no, this is not a scam. Karen is a real person and I know her and her daughters very well. Our relationship has been a bit of a rollercoaster—mostly due to the nonstop onslaught of life-threatening illnesses—but she’s a good person who needs help.

Thank you for reading.

—Robert

The numbers are slippery, but even conservative estimates suggest that thousands of books on Amazon now involve some degree of AI assistance, with certain categories (romance, sci-fi, low-content publishing) seeing particularly heavy experimentation. Kindle Direct Publishing has already implemented disclosure guidelines, which tells you everything you need to know: this is no longer hypothetical.

AI is here. It’s being used. And it’s not going away.

But the conversation around it—especially among writers—is still stuck in the wrong place.

Let’s look at the main objections.

1. “AI will put writers out of work”

Setting aside the obvious rejoinder that new technology always puts people out of work—

Yes, AI lowers the barrier to entry. It allows more people to produce readable text at scale. That will increase competition—especially at the low end of the market.

But writing has never been a scarcity game. The constraint has never been the ability to produce words—it’s the ability to produce something people actually want to read.

Readers are right to avoid prose that looks like it was written by a machine, just as they are right to avoid other kinds of bad writing. If poorly written AI-generated books flood the market, it will put a premium on:

• voice

• taste

• originality

• narrative control

In other words: the exact things that separate professionals from amateurs.

AI doesn’t eliminate the need for writers. It makes the difference between good and bad writing more obvious.

2. “AI prose isn’t good enough”

This one is… partially true.

AI writing has tells. You’ve seen them:

• Overly balanced sentences (“not only X, but also Y”)

• Generic emotional language (“a mix of hope and uncertainty”)

• Vague abstraction instead of concrete detail

• Dialogue that sounds like it passed through HR

• Excessive use of em dashes—something near and dear to my heart

It can feel smooth, but oddly hollow. Competent, but forgettable. And if you rely on AI to generate full prose, those weaknesses stack up fast.

But here’s the mistake: people assume this is the only way AI gets used. It’s like judging a word processor by the quality of its autocomplete.

AI doesn’t have to write your prose to be useful. In fact, that’s often the worst use of it.

3. “AI will stunt writers’ growth”

This is a more serious concern.

If you outsource too much thinking, you absolutely can weaken your skills. That’s true of anything—calculators, GPS, spellcheck.

Used badly, AI becomes a crutch. Used well, it becomes a multiplier.

The difference is where you apply it.

If you use AI to:

• generate plot ideas you never examine

• write scenes you don’t revise

• solve problems you don’t understand

…then yes, you’re opting out of growth.

But if you use it to:

• explore alternatives

• pressure-test story logic

• brainstorm variations

• speed up mechanical tasks

…you’re actually increasing your reps, not decreasing them.

The tool isn’t the problem. The workflow is.

4. “AI writing has no soul”

This claim is really two different arguments:

4a. AI prose feels lifeless

In other words, the lack of soul in AI writing is something we can experience.

Again, this is sometimes true. Bad AI writing can feel flat, generic, or emotionally unconvincing. But so can bad human writing.

“Soul” in fiction is usually just:

• the right balance of specificity and mythic resonance

• emotional clarity

• meaningful detail

• control of rhythm and voice

When those are missing, we call it lifeless.

That’s not a metaphysical problem. It’s a craft problem.

4b. AI doesn’t have a soul itself, therefore it cannot produce good fiction

This is the philosophical version of the claim. AI has no inner life. No memories. No desires. No dreams. All true.

But here’s the uncomfortable question:

How much of that—even with a human author—actually survives the translation into text?

A novel is not your inner life. It’s a constructed artifact made of words.

Readers don’t experience your soul directly. They experience sentences.

And those sentences are judged the same way, regardless of how they were produced:

Do they work?

Do they create emotion, tension, meaning?

Do they hold attention?

If the answer is yes, the origin becomes less relevant than people would like to admit.

The real test: output

At the end of the day, fiction has always been a results-driven medium.

Not intention. Not process. Not purity.

Output.

The proof is in the pudding.

And once you accept that, the conversation shifts from:

“Is AI allowed?”

to:

“How do I use this to improve my process, increase my output, and strengthen my prose—without breaking what makes my work good?”

There is no such thing as an “AI novel”

At least not in the way people imagine. Not yet, anyway.

There are only novels written by humans who may or may not be using AI at various points in their process.

And those points are incredibly flexible.

You can use AI to:

• generate names

• explore character backstories

• test alternate plot paths

• punch up dialogue options

• summarize your own drafts

• find inconsistencies

• speed up research

• break through blocks

None of that turns your book into “an AI novel.”

It turns you into a writer with better tools.

What happens next

Whether you like it or not, AI is coming for the fiction market.

Not in some distant future. Now.

The volume will increase. The noise will increase. The experimentation will increase.

You can either:

• refuse to adapt

• or learn how to win in that environment

Over the next few weeks, I’m going to show you practical ways to use AI in your writing process without losing your voice, your standards, or your soul.

Because the goal isn’t to replace the writer.

The goal is to make the writer more effective.

If you want to go deeper

I’ll be running a workshop on exactly this topic at BasedCon:

AI Without the Slop

We’ll cover:

• where AI actually helps

• where it quietly hurts your work

• and how to build a workflow that makes you faster and better

More details soon.

Besides being overconfident, I was also impatient. I disliked elaborate planning and reading instructions. The result of my lack of patience, humility and skill was that I wasted a lot of time doing things wrong before finally happening on the correct (or at least marginally workable) method. Still, through three years of trial and error, I become passably competent at home renovation. Shortly after my son, Willem, was born, we sold our small house for a healthy profit and bought my mother-in-law’s house. She moved into a house on my sister-in-law’s property, about five miles away.

My mother-in-law was happy to sell us her house because it meant that even though she was moving, she wouldn’t have to say goodbye to the house she had grown fond of over the previous ten years. What she didn’t realize is that even though there were no mushrooms growing in her bathroom, I had more ambitious plans for this house than for our last one. Her cabinets and countertops were dated, and I disliked the way space was divided in her house. Literally the day my mother-in-law moved out, I took a sledge hammer to her cabinets. I think she took that a little personally.

I planned to replace her carpet with wood flooring, move several walls, and repaint the entire interior, among other things. Fortunately, my job afforded me plenty of free time.

After five years at ADP, I realized I had gone as far as I could go at that company. Actually, I had realized it some time earlier, but I had worked myself into such a cushy position that I could hardly bear to leave. My main responsibility at that point was to maintain an internal website and troubleshooting database, but having gotten fairly proficient at software development, I had written programs to do most of my job for me. I “telecommuted” three days a week, which meant that on Mondays, Wednesdays and Fridays, I checked my voicemail a couple of times a day and spent the rest of the day working on whatever renovation project I was occupied with at the time. On Tuesdays and Thursdays, I drove in to the office around 10:30am, took a long lunch around noon, and left for home around 2:30pm. I spent most of the two hours or so that I was actually at my desk downloading songs from Napster.

Eventually I got bored with this routine and quit. It was strangely demoralizing to have a job where nobody realized you weren’t actually doing any work. I decided I needed more of a challenge and took a job doing web development at a small consulting company. This was at the beginning of 2000, just before the collapse of the technology sector known as the “dot com bust,” or more succinctly, the “dot bomb.” I realized not long after I started this job that my new employer wasn’t going to be in business much longer. Not comprehending the magnitude of the downturn that was taking place (and possessing, as usual, an exaggerated sense of my own ability), I decided to quit and start my own consulting company. This was, it turned out, a bad idea.

I’m lousy at sales, managing finances, dealing with customers, and pretty much every aspect of running a software consulting business other than the software part. Even in the best of markets, I would have had a hard time getting by as a sole proprietor – and this was far from the best of markets. Most people will remember 2008 as the year that the U.S. economy went to hell, but for anybody whose job depended on the Internet, the year of the collapse was 2000. In fact, 2008 was relatively uneventful for those of us in the tech sector, because tech companies had been so hesitant to hire following the 2000 collapse that even by 2008, there wasn’t much fat to trim.

I garnered a few small gigs designing websites, but I wasn’t making nearly enough to cover our expenses. I was going to have to find another job. Oddly, it was my mother who found what turned out to be my next job.

It was August 2001. Julia had just given birth to our second child, our daughter, Madeline. My mom had flown out from Michigan to help Julia with the baby, and one day she was reading our local newspaper when she came across a classified ad for a “webmaster” for the city of Tracy – which was about a twenty minute drive from Ripon. Although I had been browsing online job sites for a while, I hadn’t even bothered to check the classifieds. I mean, who advertises for a “webmaster” in the classifieds? Even in 1998, that would have seemed a little odd, and in 2001 it was downright anachronistic. In 2001, nobody was hiring webmasters at all.

I applied for the job and was one of about twenty people called in for an interview. To my surprise, I was offered the job. Now while I’m a very able web developer, it seems unlikely that I was, on paper, the best qualified candidate. There were at this point two major flavors of web development: Microsoft’s Active Server Pages (ASP) and an open source technology known as PHP. The city’s website was built using PHP, and my experience was almost entirely with ASP. Not only that, but two of the three people on the hiring committee were non-technical, and even the technical guy only asked me very basic questions. The only conclusion I could come to was that I had been hired based on my personality.

You may find this hard to believe, given what I’ve told you about my interpersonal skills, but once I make up my mind not to be an insufferable prick, it turns out that I can be downright likeable. Charismatic, even. Every place that I’ve worked (at least since my chili cheese dog epiphany), I’ve somehow managed to be the one person who gets along with almost everybody. I guess it’s my self-deprecating wit or something. Paradoxically, the less effort I put into fitting in, the more accepting people were of me. At Charlotte High, where I was desperate to find my place, I stood out like a sore thumb. But when I applied for the webmaster job, I realized that “fitting in” would only make me another forgettable candidate. To get hired, I needed to be remembered, and to be remembered, I needed to stand out. It also helped that the odds were so stacked against me that I was beyond desperation: I figured there was basically no chance of me getting the job, so I didn’t try that hard. My resignation evidently came across as confidence. Anyway, whatever it was, the hiring committee liked me, and I got the job.

Given the economic climate, I was thrilled to get the job, but unfortunately it was only part time, and I still wouldn’t be making quite enough money to meet our expenses. It became clear that I was going to have to do something to make some real money. What I did was to buy ten acres of empty land just outside of Ripon.

Now that may sound like borderline insanity, but hear me out. Thanks to my renovating my mother-in-law’s house (and a strong California real estate market), we had about $50,000 in equity. Ideally, we would have taken out a second mortgage to get us through this rough patch until I could find a job making more money, but we couldn’t refinance because I wasn’t making enough money to qualify for a mortgage. The only way to get that equity out was to sell the house.

The problem with selling our house was, of course, that we still needed a place to live. Around this time, though, we found out that couple we knew, Matt and Gail, had just bought two acres of farmland just outside of town and were planning on building a house on it. And I don’t mean they were planning on hiring contractors to come in and build a house; I mean they were actually going to build a house. Matt was a high school woodshop teacher and was pretty handy. The farmer who was selling the land had only recently found out that his orchard had at some point been subdivided into several smaller parcels, which meant that he could sell some of them off while still holding onto the bulk of his land.

My thinking was: if Matt can build a house, then so can I. This farmer probably has plenty of cash, and I might be able to convince him to finance the property himself if I make him a full-price offer with good terms. I’ll make a low down payment, cash in my 401(k), and use the money to build a small house. We’ll find somewhere cheap to rent while we’re building and then move into the house when it’s finished. Hopefully by the time the house is done, I’ll have found a higher-paying job, and we can refinance, pay off the balance of what we owe to the farmer, and maybe even have some cash left over. I asked Matt if the guy had any parcels left.

It turned out that all the two-acre parcels were sold, but he did have one ten-acre parcel left – for which he wanted $300,000. This was actually a very reasonable price, but I had hoped to offer the farmer a ten percent down payment, and I couldn’t afford to give him $30,000 and still build a house that met the county building codes. Some tweaking of my plan was in order.

I approached my father-in-law, who was at this time living in a single room in my brother-in-law’s house. I knew he was looking for another place to live, and I suggested that he buy a mobile home and move it onto the ten acre parcel we were looking to buy (after we bought it, of course). The deal would be this: he would pay for the mobile home, and Julia and I would live in it for a year or so while we built our house. After that, we would move into the new house, and he would move into the mobile home. In exchange for letting us live in the mobile home, we would let him live rent-free on our property for as long as he wanted.

My father-in-law agreed to this, and I made a full-price offer to the farmer, with a $30,000 down payment and the rest due in two years. He accepted the offer. There was a bit of a delay between the closing on our house and getting the mobile home set up on our new property, so we spent about seven months living in a miserable two bedroom apartment. Other than that little hiccup, though, it looked like my plan was actually going to work: we’d be able to get by for the next two years, after which we’d refinance, pay off the farmer, and be sitting pretty.

My mother-in-law, not the most adventurous person (and probably still a little put out by my sledge-hammering her cabinets, not to mention selling her house at a healthy profit only two years later), was dead-set against the plan. In retrospect, she made some good points (“what if the real estate market collapses?”), but to me her protests just seemed like gratuitous negativity. My own mother wasn’t much more encouraging. But mothers worry. That’s what they do.

My father-in-law was on board with the plan, and my own father was surprisingly supportive. My parents came out to visit while we were negotiating the deal, and while my mother and mother-in-law shook their heads in disapproval (and Julia stood by uncertainly), my dad looked about the property in wonderment, saying, “Boy, this will be really great if it works out.” My father had by this time adopted a sort of almost Zen-like sheepishness, becoming so aggressively amenable that I barely recognized him. It was like he had sworn off criticism or negativity of any kind: he greeted everything and everyone with a sort of detached bemusement. It frankly became downright irritating at times: I would ask him for some practical advice – about, say, the sprinkler system I was putting in – and he would reply, “Oh, I’m certainly no expert on that.” He didn’t seem to understand that not only did I not really need advice on installing a sprinkler system; it actually pained me to ask anyone (let alone him!) for advice. I was simply trying to communicate with him – about sprinklers, if that’s the only common ground we could find. Whereas my attempts to relate to my father had once been rebuffed by his sarcastic reprisals, now I was trying to engage a mere shadow of a person.

Still, it was a nice change to actually hear my father offer some words of encouragement regarding my plan to build a house on ten acres of farmland. But the father of my childhood loomed larger in my mind, and it was ultimately the impression left by that father which nearly derailed the plan. Thanks to my dad, I had developed a deep-seated intolerance for assholes in positions of authority, a trait that still tended to get me in trouble.

My new boss, the head of the IT department for the City of Tracy, was a pony-tailed, marginally competent sad-sack, the kind of guy who spent his life trying to live up to his memories of Woodstock, even though he was only in 3rd grade at the time. Stoner, as I’ll call him, was the worst boss I ever had.

To be continued…